19 October 2010

Mesh: Dead on arrival

There are a lot of people anxiously awaiting the arrival of mesh importing on SL. They think it'll be neat stuff and make the world a better, richer, more interesting place. I think it's dead on arrival.

For those of you who don't know what that is (yes, there are probably some yet), meshes are a way to build objects with any shape desired in a 3D modeling program like Maya or Blender and have them appear in SL just as they were designed. A mesh describes an object as a series of vertices, with lines and surfaces connecting them. Any object can be modeled this way; how well is a matter of how many vertices are used.

SL has actually had a form of meshes for quite some time. A sculpt is a mesh, but limited in the level of detail by the fact that it can have at most 1024 vertices. That's enough for quite a lot of creativity, but it's not enough for many things. Sculpts are usually used to make one part of a larger item, in greater detail than can be done just with primitive objects like boxes and spheres and cylinders.

A mesh doesn't have that limitation. One demonstration I've seen is of an Audi R8 in one 8192-vertex mesh. It looked quite impressive. That's the kind of eye candy that Linden Lab is hoping creators will make, and that users will demand. People have been demanding it for years and eagerly awaiting it ever since LL announced they'd do it.

There's only one problem: Mesh is dead on arrival. As currently planned, it's got far too many things wrong with it to be useful to content creators.

The first hurdle is that it requires massive changes to the viewer's scene rendering code (the part of the viewer that draws the picture shown on your computer screen). These changes, practically speaking, are only practical in LL's Viewer 2 or third party viewers based on it. This runs smack dab head-on into the basic fact that the SL user community hates the Viewer 2 user interface with a passion normally reserved for a kid hating Brussels sprouts. LL's been trying hard to get users to Viewer 2 ever since it was introduced early this year. The latest figure I've heard is that it's now used for 20% of logins to SL. Where are the users? Either still on 1.23 (which Viewer 2 just passed in popularity for the first time last week), or else third party viewers. Before Emeraldgate, Emerald had 60% of the userbase all by itself. Since then, the users have scattered, with most moving to Phoenix, but a large number are using Imprudence, Emergence (which is deprecated by the developer, who recommends that people use Phoenix instead), and Ascent. All of those are based on the version 1 codebase, which will not show meshes.

All of that means there's a chicken and egg problem. Content creators aren't going to bother with meshes unless there's significant demand, and a significant number of users who can actually see the content they'll create. Users aren't going to switch to Viewer 2 just for meshes that aren't being created, no matter how much LL might think otherwise. Oz Linden said that he expected mesh to drive users to Viewer 2 at a recent third party viewer developer meeting. I think he's full of prunes. That's wishful thinking. Mesh won't overcome users' hatred for the Viewer 2 user interface.

There's another problem with meshes for content creators: Oskar Linden announced at yesterday's mesh office hour that no third-party viewer would be permitted to use LL's mesh upload code in their viewer. It seems that the upload code uses part of the Havok physics library that LL licenses from Intel. That library is very much not open source, and LL is not permitted to distribute it except as part of their own program. This means that a content creator will have to exit his viewer of choice (almost none of them are using Viewer 2, because they can't afford to take the weeks needed to become proficient in the UI), start up Viewer 2, upload the mesh, then sign out and back in with his regular viewer. To a big content creator, time is money. If using a mesh takes up that much time and effort, they just won't bother.

Meshes are also considerably less useful than they might appear to be to clothing designers, because they cannot be made flexible. A piece of rigid clothing isn't very comfortable or realistic.

Finally, the cost of using meshes, as things currently stand, is astronomical. There are only two things in SL that cost real money: land area and primitive usage. That's because both translate directly into load on the server computers that run Second Life. Meshes also add to that load, so they need to count as well. Just how much they cost isn't known outside the group within LL that is doing the project (and maybe by nobody, really, but Andrew Linden). This cost is being expressed as an equivalent number of primitives, and will count against the primitive limit for a parcel of land. The problem is that that cost is being set high  enough that, in many cases, a mesh object is more expensive than the same object built up out of primitives. Yes, the mesh object looks a little bit nicer, but if the price is double that of the primitive-based object, why bother?

As I said above, Linden Lab has a chicken-and-egg problem with mesh. The rest of the problems with it pretty much guarantee that people aren't going to make the leaps needed to exploit them. Despite LL's best hopes for meshes pushing people to Viewer 2, it just ain't gonna work that way. Mesh is dead on arrival.

41 comments:

  1. I have to agree. In my group blogs, some time ago I stated, "I'll believe mesh when I see it. A company that can't get sculpties to rez isn't likely to do any better with meshes."

    This article is very insightful in the "chicken or the egg" situation. If 80% of the viewers on SL can't see mesh constructs... it's ridiculous for merchants to even bother with them. Linden Lab may be hoping merchants will do so anyway, forcing people to v2 in order to see all the nifty new mesh creations. Whether that actually happens or not remains to be seen.

    In all of this my one, single, pervading question has to be this: Why doesn't Linden Lab wise up and put the v1.x interface on v2.x... and eliminate the problems users have with the v2 product? In all my career, I have seen few companies as stubborn as Linden Lab. They don't seem to be able to say "Maybe we were wrong"... much less change their course to the path of common sense.

    If customers want the v2 product but hate the v2 interface... put the old interface in v2. That's kind of one of those duh things, isn't it?

    I think eventually mesh will get off the ground... if Linden Lab survives that long. I think it will be far more likely to get off the ground on the OpenSim constructs using third party viewers. The TPV creators are already finding a way around the Havok limitation. What, Linden Lab thinks we have no skillz of our own? ; )

    ReplyDelete
  2. LL may claim that the 1.x code can't do mesh, but they neglect the fact that TPV devs can backport the mesh code, and there are already plans to do so.

    ReplyDelete
  3. Wayfinder: The problem with putting the old interface on the v2 viewer is that there's a lot of code down deep that makes assumptions about the way v2 does things. It'd be almost a total rewrite. They spend a few million bucks on developing v2, and spending a couple million more on another rewrite would get the board of directors interested - and not in a good way. Some heads would roll. As for finding a way around the Havok limitation: I sure hope we can in Phoenix...

    Zauber: There are no doubt other TPV teams looking at adding mesh to the 1.x codebase. I know we on the Phoenix team are not; we're concentrating on making the current viewer stable so we can concentrate on putting out a v2 viewer with a UI that doesn't suck. We're just beginning to look at that now, since we've spent the last 6 weeks just trying to get something good and stable for people to use while we go off and spend 6 months on v2.

    ReplyDelete
  4. Tonya, I first wish to say there are tens of thousands of people (hundreds of thousands?) who really appreciate the work you folks do on TPVs. Without that work we'd all be stuck towing LL's line. You folks don't get any reward but gratitude and not a whole lot of that. So... thanks.

    Regarding changing v2 to a decent interface-- I don't think it would be all that difficult. All people are really asking for is that they return the previous menu structure so we know what the dickens is going on where... and that they exercise some kind of common sense when adding anything else to the system.

    Yes, some things would take a re-write, such as going back to the original IM layout and getting rid of those ridiculous system message black boxes that keep popping up and obscuring the screen. But those things are few. For the most part, people simply want the old menus back, we want to be able to find the tools we need where we need them, and we want Linden Lab to stop stabbing their long-time users in the back in their effort to attract newbies (a failed effort at that).

    I know I'm probably preaching to the choir, but all they've done is created an interface that no one wants (not newbs or current users)... and they've alienated their clientele in the process. Whatever it costs them at this point, I think their smartest move might be to backtrack a little and work very hard at keeping the customers they've got. Because if they don't start shaping up, from what I've seen over the last two years they're likely to lose even those (and indeed, they already are).

    ReplyDelete
  5. The biggest challenge associated with Mesh is demand. When LL was pursuing the business space, mesh was a more important feature. After all, there are billions of dollars worth of 3D business assets in RW CAD programs.

    If you have a product that provides a solid ROI for businesses that want to do virtual prototyping, collaboration, training or demos, being able to import mesh objects/structures is a highly attractive feature. Now that LL has seemingly abandoned their interest in promoting SL for business, the reason for making a huge investment in mesh baffles me.

    Having said this, I still hope SL finds a way to pull together a viable business offering for the SMB (Small to Medium-size Business) market, who are best positioned to benefit from what SL could offer them. This type of offering should include mesh as well as a channel program to sell, train on and better support it.

    ReplyDelete
  6. If you think encroachment is bad now I invite you to go to the mesh preview and look at the gigantic crap people are rezzing out that cover multiple regions and interfere with movement and cause crashes.

    LL hasn't the human resources to deal with the magnitude of what is to come.

    ReplyDelete
  7. The part about TPV's not being able to upload Mesh isn't 100% accurate. Yes we won't be able to use their Convex Decomposition library which uses Havok to map the physics on meshes, however, they've provided a reference stub in hopes that OS devs can find a solution they won't dedicate their time & money to themselves. There's already devs talking about adapting Bullet to their Convex Decomposition stub to generate the physics details for meshes. So Mesh in TPV's will be possible, just not immediately.

    ReplyDelete
  8. Yes, after the Second Coming of Philip, the Mesh God seemed to many to be the Saviour of SL. I am, and always have been, sceptical, a fact that has got the disciples baying like mad wolves on my blog.
    Now, with no emote possible on an alternative avatar, no flexi for hair or clothes, and problems uploading large files, I think that it will be a long time coming.
    The Infuriating Viewer is so hated I doubt it will ever be widely accepted and your argument adds credibly to the debate.
    Great post.

    ReplyDelete
  9. Ann: Oh dear. Huge meshes that overlap sim boundaries? Who thought that letting that happen was a bright idea? I do know that the mesh sim code does allow prims up to 64 meters as well, and presumably raises the 53-meter linkset size limit as well. Those are both good things. Still, you're correct in that LL doesn't have the people to deal with supporting this stuff, simply because they don't have the people to deal with many, many things they should be dealing with now.

    Ron: I do hope the situation with mesh upload isn't as dire as it seemed at first. LL really isn't out to kill off TPVs except by competing with them, and that limitation really fueled the conspiracy theorists.

    Soror: I didn't know that they weren't allowing emotes on alternate avatars...Hoo boy. Yet another limitation that will knock out a good chunk of the market. What's left? My AnthroXtacy tigress would be a lot less interesting if she didn't blink and her jaw didn't move with speech.

    ReplyDelete
  10. I should also note here that one developer, at least, has separate the mesh renderer out from the rest of the code, and others are looking to adapt it to v1-based viewers. I don't know, if someone else works up a patch, whether it will be added to Phoenix (just as I don't know the same thing for display names), but I'm sure someone will. If that is indeed possible, it will change the likelihood that mesh will be used by more than a very tiny fraction of users by quite a bit.

    ReplyDelete
  11. "Ron: I do hope the situation with mesh upload isn't as dire as it seemed at first. LL really isn't out to kill off TPVs except by competing with them, and that limitation really fueled the conspiracy theorists."

    The question here is: are they "competing" by taking unfair advantage of internal structures... structures not available to TPVs?

    This of course is a side-thread concept. I've always felt LL shot themselves in the foot by releasing client code to open-source. It was just an insanely goofy business move-- good for TPVs and OpenSim, but potentially fatal to Second Life. Very bad business decision.

    But having done so, are they now trying to take back what they've given? It seems more and more LL is trying to shut out TPVs... which is (possibly) one of the reasons they are so stressing v2.x. I'm not a TPV dev so I don't know: are they making all 2.x code available to the public? (ie, are they releasing 2.x source code via open source?). If not, that's a big red flag. If they are... then the second question would be does their code use licensed library routines not available to TPV devs?

    The question also is this: in releasing their source code to the public, did LL intend the development of TPVs... or that their customers would help them to debug a very buggy viewer 1.x? If they didn't want TPVs developed, why didn't they release a restriction license along with the open source code?

    What I see here is a company with a history several years of really ludicrous and short-sighted management decisions, some of which were extremely harmful to both customers and company (OpenSpace / Homestead fiasco) and others of which are just plain abusive-- repeatedly. Some are questionably suicidal from a business standpoint. Very, very few were actually beneficial, to either customer or company... as evidenced by the current state of Second Life.

    It's pretty sad when it seems we customers know more about how SL should be operated than management itself. There are (at least) two aspects to managing SL: the tech (keeping the grid running, coding, etc) and service (keeping customers happy and basic understanding what SL is all about). Tech seems to be constantly hampered by service (MORE TOYS!) and service is clearly clueless as to what customers really want.

    No arrogance or exaggeration intended, I think I could manage SL better without even breaking a sweat. I believe I'm not the only one of experience SL user/professionals that could reasonably make that claim. Imo, it would be difficult to manage it worse without total collapse.

    ReplyDelete
  12. Tonya, the logic to determine the prim cost of meshes is very much work in progress. To reach the right balance between detail and rendering cost at least 3 revisions and a couple of tests were made during the closed beta. This issue was amply discussed during meetings in the closed beta. The Lindens made it clear that they want meshes to be used by SL users but they need to take into consideration their streaming cost, else the result would be increased lag. What they want to avoid is to repeat the mistake made with sculpts when it was determined that their cost had to be 1 prim.

    The current revision is costly in terms of prims but the Lindens are aware this is a problem and they are working on it. A couple of interesting observations have been made by beta testers on this topic and a more recent comment I read suggests that a further revision of the cost logic bears better numbers. I believe the Lindens will find the right balance. While meshes will never be as cheap to use as sculpts for the reasons mentioned above, they will bring benefits for all those objects that will need detail. Which, by the way, should put an end to the diatribe that meshes will be the end of prims.

    ReplyDelete
  13. really, looks like som folks are paid to throw c**p @ LL. Since i read a lot of "just to say something" stuff, i won't argue anything. to each of his/her own.

    regards

    ReplyDelete
  14. Rock, I see a lot of folks defend Linden Lab when in my experience, such defense is unwarranted. In all that I read in the above posts, the discussion is factual and pertinent. Like you, not arguing, just making a point. If Linden Lab doesn't want c**p thrown at 'em... maybe they should stop throwing c**ppy management decisions at us and start operating the grid in a manner that's beneficial for their customers (for a change).

    They've been so busy the last few years putting profit first and focusing on their bottom line, they seem to have forgotten WE ARE their bottom line. Tick off customers... profit ceases.

    ReplyDelete
  15. Wayfinder: Thank you. You've prompted me to comment on the whole open source viewer argument. I'll post an entry today about that.

    Indigo: I do hope that LL will cost meshes more appropriately. Even if having one means that you can have other things on your 512 sqm parcel, though, the other problems are still quite real and enough to sink mesh usage. Essentially, LL has made all the wrong choices to make meshes anything more than just a curiosity. Can those choices be reversed? Probably. Will they? Given LL's recent history, I'm not all all certain.

    ReplyDelete
  16. Tonya, I disagree. None of the points you listed are real show stoppers in my opinion. You say that the Phoenix team will develop a version based on a V2 code base with a different interface. I believe the Imprudence team is doing the same. Thus, apparently there is an issue on how long it will take for TPVs teams to catch up, not that it is a major feat to the point that it is not feasible. This may slow down the adoption of meshes but it's hardly a show stopper.

    The issue with uploading meshes is related to licensing problems, as you point out, but there are talks about using open-source libraries already. As for clothes, the vast majority of cloth parts based on prims are not flexible, so how is that different from today? Incidentally, meshes can't be made flexible for very good reasons.

    I also think that you underestimate the impact meshes will have on content production and user demand. Mesh-based models are without any doubt better looking and more detailed than anything one can do with prims and sculpts. Even the texturing quality improves a lot. Anyone who takes building seriously will appreciate the fact that they can use more sophisticated tools and the improved workflow. And, most of all, will appreciate having copies of their models on their hard-disk without any effort. It just takes a look on the beta grid to see how much interest meshes cause. The Flickr group dedicated to meshes that I created is a success: 235 photos, 81 members in less than a month. The most viewed photo, the Oak Tree made by Timmi Allen, has over 10,200 views.

    In my opinion, TPVs have to hurry up to implement mesh rendering into their viewers because there will be a big demand.

    ReplyDelete
  17. Hey Wayfinder-

    I don't think that LL is trying to sink TPV. It, at this point would be foolish to do so. There is no magic bullet viewer solution. At this point only a fool would think otherwise. TPV takes some of the pressure off LL to create such an impossible beast. That being said however, LL has a real problem. They spent a significant amount of R&D money on V2 and it not only didn't catch on with current residents, it didn't do what it was SUPPOSED to do- which was attract and retain new ones. The best the Lab can do right now is say that there's 20% using v2, and that it's now surpassed 1.23 usage. This number is almost laughable in how meaningless it is. Many people have left SL entirely, dropping the 1.23 user count. New users who don't know any better *have no options* but to use 2.0- it's a while before they find out that there are any other options to begin with. But they are desperate to make v2 look like a silk purse rather than a sow's ear- if they *force* a 2.0 solution by not sharing fully with tpv, they get to pad their numbers and the v2 project looks better on paper.

    -AK

    ReplyDelete
  18. @DGD: Well, with most companies I'd agree with you. It makes no sense to battle with the TPVs, so why would they do it? It would be foolish to do so, downright suicidal.

    But in this case, we're talking Linden Lab. This is a company that amazingly seems to wallow in senseless, suicidal decisions. The OpenSpace/Homestead debacle surely wasn't sensible. Alienating the entire EDU community-- some 10% of their population and 8% of their sims-- was not sensible (downright stupid if ya ask me). v2.x was/is a fiasco. They've been warned about the potential problems with Display.Names even before it hits the grid, yet they're forging on ahead with that one. This company seems bent on a course of self-destruction.

    As far as "no magic bullet" cure, I agree. Any "cure" to the viewer is going to take some common sense, dedication, and a bit of work. That doesn't mean it can't be done. I really don't care how much they have invested in v2.x... if it doesn't work, the customers don't like it, and it's been rejected by 80% of the population-- recognize it as a bad investment, salvage what "feature" code they can, scrap the whole project, fix v1.23 (which is what they should have done in the first place) and release a sensible product to the grid. Either that, or change the v2.x menu layout and performance so that it more closely resembles the old viewer (and every other TPV on the market) and admit they were wrong to try and mess with that.

    Or even better... they could bring back the 1.23 menus in an improved format-- as has been proved successful by the TPVs.

    I'm an old business consultant. "That can't be done" is in my mind, the equivalent of "I want to be fired because I'm too lazy to do my job or too incompetent to figure out the right way to do it." The decisions that have been made by Linden Lab over the past two years-- that's a reason to fire management and bring in people who know how to run a business.

    ReplyDelete
  19. Follow up: not that they *would* fire management, because most likely the owners are the managers. But any other company out there, man, heads would roll. M. Linden would have been the first in a line of "who is responsible for this mess?"

    Not that I'd fire people. Experience counts for something. Far better to bring in an Iococca and start teaching folks how to do their jobs the right way.

    ReplyDelete
  20. Indigo: The problem with v2 TPVs rescuing mesh is that doing right with version 2 is a 6-month project at the least. Yes, TPV makers are doing it, because we have little choice (see the post after this one). That doesn't mean it will happen quickly.

    As for flexible meshes being useless for clothing, don't tell that to folks who make clothing; they'd use them if it were possible. (Mesh hair, anyone? Skirts? Furry attachments?) Whether there are good reasons for it or not, that doesn't change the fact that having them be useless for clothing is going to prevent one entire class of content creators from using them.

    The improved workflow might be a big win...but Viewer 2 is just as big a loss, if not bigger, so it's a wash at best. If a content creator has to switch viewers to upload something, he's going to find a way to get around it. Yes, it's possible that an open source library will allow uploading meshes. If so, that's great; if not, it's another nail in mesh's coffin.

    What it boils down to is that you can't have mesh without Viewer 2, and Viewer 2 is bad enough that people won't switch to it, no matter how good meshes are - and meshes have other problems that make them not as good as advertised.

    When TPVs that are actually usable for folks on the Viewer 2 codebase are available, the equation might change, but right now, content creators aren't going to make stuff that only works for 20% of the users, and users aren't going to switch for nonexistent content. No amount of wishful thinking is going to change that.

    ReplyDelete
  21. A pity there's a six month effort ahead for TPV developers to get ported to the Viewer 2 codebase. Shame there wasn't any way for them to start six months ago. Oh, wait.

    Okay, well, at least Lindens could have hinted that the Viewer 2 codebase was the way to go. Or if not the Lindens, then other TPV developers. Oh wait.

    Let's cut to the chase: In a few months--probably a few fewer than six--folks will have to choose between their beloved bouncing bewbies and not bumping into bounding boxes their first-generation viewers can't see. Eventually users will quit stubbing their toes, just because it feels so much better when they don't.

    Maybe some developers will, too.

    ReplyDelete
  22. qie, your argument assumes that people will indeed produce meshes to any significant degree in the absence of a large userbase equipped to use them. It is precisely this argument that I think is wrong. I don't think content creators will produce significant numbers of mesh-based of products until more than a small minority of users can see and enjoy and *buy* them.

    As for the "you should have started 6 months ago" argument, you seem to have forgotten that LL said, when Viewer 2 was announced, that it would exist alongside 1.x. That clock should only be started when LL announced that 1.x would indeed be end-of-lifed, and that was much more recent.

    It's a shame that Viewer 2 was so sucktastic. If it hadn't been, then people would be using it, and TPVs would be working on it, and we'd be a lot farther down the road.

    See? I can play the 20-20 hindsight game too.

    ReplyDelete
  23. As for "folks will have to choose between their beloved bouncing bewbies and not bumping into bounding boxes their first-generation viewers can't see", there's much more than bouncing breasts that people would lose when moving to Viewer 2. Viewer 2 is unusable for prolific content creators, for example, because you can't batch edit permissions any more, greatly slowing down the content publishing process.

    Sounds like you love Brussels sprouts and insist that everyone else must live them too, or else. But this argument properly belongs on my next post, not this one.

    ReplyDelete
  24. Tonya: "What it boils down to is that you can't have mesh without Viewer 2, and Viewer 2 is bad enough that people won't switch to it, no matter how good meshes are."

    I'll second that opinion. I'm one of them. I absolutely refuse to use v2. I gave it a fair chance, twice. In both cases I wiped it off my system within 2 hours. What a piece of carp. (typo)

    As for whether merchants will create things with mesh or not... I'm sure they will. Will people buy them? Some will. Whether it will be widely accepted or not, that's another thing entirely.

    I know I wouldn't buy anything that 80% of my visitors can't even see. Non-rezzing sculpties are bad enough.

    Let me be frank: my group started exploring Inworldz last April. There is a joke there: "It's buggy, it's laggy, it's full of problems, but at least it's not Linden Lab."

    While over the past two years we shut down 7 sims on Second Life due to abusive Linden Lab decisions, we have purchased FOURTEEN sims on Inworldz (with 45,000 prims each!)... and we have plans for more. Even without physics, with bugs galore, with the script system not fully developed, our group can still function there. What's more, we are enjoying ourselves there much more than we have in years... and our builds (without the arbitrary and totally unnecessary building restrictions of SL) are far beyond anything we could have developed on Second Life. It's like a new breath of fresh air.

    That's pretty much a slap in the face to Linden Lab, that people would prefer a technically inferior platform to dealing with their company attitude, restrictions and prices.

    I realize of course, there are some people who like v2. (I have to believe they are few and far between, and mostly newbies who don't know any better.) Some prefer Phoenix, some Hippo, some Imprudence, etc. Some people like Inworldz, some hate it. The same can be said for Second Life. Diversity is the spice of life. But the reality is 80% of the SL userbase is NOT using Linden Lab's viewer... and they are losing thousands of customers. Second Life sim count is actually decreasing rather than increasing. And their managers are actually getting PAID?

    You'd think they'd get a clue, y'know?


    @Qie: It's easy to say "should have started 6 months ago". Started with what? No one know how LL would implement meshes, what form they would use, whether they would go with industry standard or come out with some cockeyed LL proprietary standard, etc. Sure, TPVs could have started on mesh implementation months ago... foregoing essential work on foundation code... and then had to scrap all that work when LL came out and said "Ok, here's what we did..."

    That would have been a total waste of time and effort. So what if it takes TPVs 6 months to get meshes together. I'd wager it'll take LL that long to work out the bugs... or longer. ; )

    ReplyDelete
  25. If mine is "20/20 hindsight", consider what must then be the amazing foresight of those TPVs who *did* base on the Viewer 2 code.

    Yeah, Viewer 2 sucked and continues to have deep UI design flaws still -- enough so that the "chicken and egg" dilemma indeed prevented creation of all that much MOAP, for example. But one doesn't bump into MOAP.

    @Wayfinder: Q: "Started with what?" A: Started porting their specific features to the Viewer 2 codebase, even if there weren't users to justify releasing the result--exactly the task ahead of them now. Arguably some of that would have needed rework as Viewer 2's UI evolved, and indeed that's a problem going forward, too. But now they'd have Mesh out of the box, along with MOAP, fast shadows, and everything else (past and future) not yet backported to the 1.x base.

    The work needed to port TPV-specific features forward to 2.x might have precluded some backporting of V2 features, or delayed introduction of some new feature; that was a decision TPV developers had to make. Some TPVs favored one path, some the other.

    But that's all sunk cost now. Going forward, hopes that Viewer 2 suckage will prevent significant Mesh adoption is wishful thinking.

    ReplyDelete
  26. @Qie: I wasn't aware that Linden Lab released pre-production information on how they were implementing mesh (did they?). To my knowledge, they just kind of sprung it on us recently (Hey! We've almost got mesh ready) and didn't let anyone know the specifics of what/how/when. (I may be misinformed; if such is the case lemme know.) Such would have been necessary for the TPVs to be able to plan any kind of mesh implementation... because LL is known for pulling proprietary "walled garden" coding and concepts. For all TPVs knew, LL may have decided to implement mesh, but with very narrow parameters.

    I also am not aware that *any* TPVs are basing on v2 code. From what I've seen, they're pretty much thumbing their noses (rightfully) at the whole v2 situation.

    I don't think "wishful thinking" has anything to do with this. I think "reality check" is more in line with what's going on. The point that's been made (valid or not) is that with a 20% user rate on v2 and 80% of the platform absolutely refusing to use that viewer, mesh may take a while longer to implement that Linden Lab is expecting... if at all.

    It's hard to predict the future of course, and while looking ahead is always wise, debates about what is to come is pretty much a waste of time. I'll agree with you that likely LL will attempt to implement meshes just like they're attempting to implement v2. Whether that actually works or not... or whether meshes will indeed be "dead on arrival"... is anyone's guess. I do think we can safely assume that meshes will not be the salvation toy Linden Lab seems to think it will be.

    Myself, I'm questioning whether SL will even be around a year from now (at least in its current form). Given that concept, the idea of meshes becomes one of far less significance.

    ReplyDelete
  27. qie: "hopes that Viewer 2 suckage will prevent significant Mesh adoption"? "Hopes"?! Methinks you have an agenda here. I would like to have meshes in Phoenix. It ain't happening. It won't happen till Firestorm. I bear mesh no ill will. I think LL is screwing it up, in all the ways that I described, and wish they weren't. If it weren't for that, mesh adoption would be much higher much faster, and we'd all be better off.

    Wanderer: There are at least two viewers I know of based on Viewer 2, Catznip and the current RLV. Those are significant because the work to do RLV in the Viewer 2 codebase has already been done, and that will shorten the TPV cycle quite a bit.

    I'm not as pessimistic about the future of SL as you are, but OTOH I did register an account on Inworldz this morning...

    ReplyDelete
  28. It's hardly been a secret that Mesh was coming, and there was enough information about how it was to work for there to be an OpenSim implementation within 24 hours of its appearance on Aditi. But unlike OpenSim developers, TPV teams don't need to care about all that because they get Mesh for zero effort if based on the V2 code -- and any other implementation approach is as near to intractable as makes no difference.

    There are TPVs based on v2. Kirsten's and RLV are in the directory; Catznip fixes some especially annoying 2.x UI problems that still persist in Snowstorm. There may well be others with which I'm not familiar.

    It seems that saying anything about the past is "hindsight" and speaking of the future is "a waste of time." That narrows the universe of permissible discourse to the fleeting present. It's raining here now, but dare I presume to take an umbrella?

    ReplyDelete
  29. qie, the reason I pointed out that your argument was 20/20 hindsight was that you took TPV developers to task for not starting on v2-based viewers as soon as v2 code was released. That's neither fair nor useful. Regardless of how we got where we are, we're here now and must deal with the world as it is. Perhaps we should have started sooner. Perhaps, also, LL should have outsourced Viewer 2 development to someone who actually knew what using SL was like and then rode herd on them to enforce a clean separateion of UI and backend layers. None of that matters now, however. We're here where we are, and we must deal with that.

    My point is that the Viewer 2 features - and, FWIW, I consider media on a prim to be a massive security hole, and refuse to use it myself - aren't enough to overcome the fact that users hate the Viewer 2 UI. This means that adoption of Viewer 2-only features will be minimal at best.

    ReplyDelete
  30. Qie: "It seems that saying anything about the past is "hindsight" and speaking of the future is "a waste of time." That narrows the universe of permissible discourse to the fleeting present."

    That's just being argumentative. No one has said that hindsight is worthless, nor that speaking of the future is a waste of time. That's twisting our words. What we said is that using "hindsight" to try and say people should have predicted the future is just a little piddly as a standpoint. Where were your future predictions 6 months ago Qie?

    I'm not seeing the point you're making Qie (is there one?). Saying "shoulda woulda oughta mighta" really doesn't serve any purpose I can see. That's what Tonya is stating. Let me explain (and no disrespect intended).

    Can you point to a blog post you made 6 months ago, recommending to the TPV people to implement Linden Lab Mesh as quickly as possible? If not, then I would offer you were just as lacking in "foresight" as anyone else. Perhaps the reason for that is that-- like with everyone else-- there was no trustworthy foresight to be gained at that time. It's not like Linden Lab has been consistent and reliable over the last two years. Their activities have been fairly unpredictable... and quite abusive.

    Even if they had been somewhat predictable in such things (which imo they weren't) that is based on the assumption that we agree with the opinion that implementing Linden Lab mesh concepts in TPV was the way to go... or that we agree with the assertion that there was enough information forthcoming from Linden Lab to do so properly (I still assert there was not).

    I believe that any work done by the TPV folks to implement meshes had about an 80% chance of proving out a total waste of time once LL finally revealed what/how they were doing (if they have even done so yet. Have they? Do we even yet know exactly how meshes are implemented on SL, or is a lot of this guesswork? Has LL released open source Mesh code?).
    (cont)

    ReplyDelete
  31. (cont)
    You have to understand that over the last 6 months, Linden Lab has thrown some severe monkey wrenches into the grid / TPV concept that has kept everyone hopping with updates and compliance issues. The TPV folks have been busy (as was the original intent) with debugging seriously faulty LL viewer code, in order to get the foundation to work right. So where in there were they supposed to have the time to put together a "might/maybe" mesh system that might-maybe work once LL released their code (IF they released their code)?

    I will point out, Qie, that even as we speak, Linden Lab has not officially released Mesh to the main grid yet. It's in testing... and people are finding a LOT of problems with it... primarily in effect on the grid as a social and logical structure. So it seems to me that even Linden Lab hasn't been exactly aware of the best way to release mesh to the grid. Yet TPV people were supposed to figure this out 6 months ago and start working on it themselves? I'm afraid I find that premise invalid.

    Further, how could anyone even be sure Linden Lab would follow through with mesh? Linden Lab has a long history Qie, of making announcements and promises and not following through. When they announced quite some time ago that "in the future they would be considering implementing mesh on the grid"... no one that I know of really took them seriously. Most folks had a "We'll believe it when we see it" attitude.

    So yes Qui, 20/20 hindsight and "you shoulda done this" based on that 20/20 hindsight is a very easy game to play. It's another entirely to try to actually predict the whimsical nature of future Linden Lab decisions and base project goals on the unpredictability of that company.

    About the only "predicting future SL" game I've played for the past two years is a rather constant "watch your butt and wait for the next dozen shoes to fall". I've done so with a fair degree of accuracy... but even I'm rather stumped at what LL is up to these days. This is a company that is seemingly without logical intent, and that regularly makes blatantly suicidal decisions (their kicking the EDU community in the gnards is just astoundingly bad business). So no, I'll have to believe that trying to predict the future of meshes 6 months ago... before LL themselves announced their actual implementation... would have been an exercise in futility and a major waste of time, especially when other things required priority attention. I don't see the TPV folks being AT ALL short-sighted here.

    ReplyDelete
  32. Tonya: "Perhaps, also, LL should have outsourced Viewer 2 development to someone who actually knew what using SL was like and then rode herd on them to enforce a clean separateion of UI and backend layers. None of that matters now, however. We're here where we are, and we must deal with that."

    I have to agree. I have been openly critical of the whole v2 debacle. Linden Lab certainly should have known better than to outsource design to a company that did not consist of people with SL expertise, but that was their decision.

    It is very likely they themselves provided the specs to that company... which meant of course that they themselves don't have a clue as to what their customers want (which comes as a surprise to no one).

    However, I have never focused on that boneheaded move, other than to say "Realize you've made a bone headed decision and fix it." The problem with outsourcing is as you say, past, and was the first mistake. That mistake was compounded when 88% of the grid rejected v2 and Linden Lab, instead of scrapping the project and starting over (doing it right this time) continued to foist that monstrosity on the public... before it was even finished.

    That changed from "should have" to "should be"... and was their real mistake (as it continues to be). I still assert that until someone at LL gets a clue and decides to majorly re-design the v2 UI, that project will continue to be a thorn in their side. Until of course, LL takes the very predictable course of forcing people to use v2 (which I'm expecting any time) by removing prior viewers from their site and redacting them from login ability.

    Oh, I can hardly wait for that one.

    ReplyDelete
  33. Tonya, I don't have enough information to judge whether it will take 6 months to develop a TPV based on the V2 code base, but if I had to make a guess I'd say that you are probably a bit pessimistic about it and that we are likely to see a compliant TPV before 6 month. Anyway, I don't mean to dispute your judgement, you are in a better position than me to evaluate the matter.

    As for clothes, you forget a detail: currently cloth parts are made both with flexible and non-flexible prims. Well, you can do exactly the same with meshes. Cloth makers can make parts with meshes and add flexible prims where needed. So, this is a non-issue.

    Linden devs who work on the mesh project are the first supporters of the technology. Qarl was probably the most enthusiastic of all, a major driving force to introduce meshes to SL, and it was a real pity that LL laid him off. Everyone wishes to have the best technology ported into SL, including them. That said, having listened to their explanations both in the closed and open beta meetings, I think Lindens have made a wise choice not to introduce some features such as flexible or double-faced meshes because the drawbacks would be severe for the system.

    Bottom line: will mesh adoption be slow? Sure. Are meshes DOA? No way. I still don't see a convincing argument to say that.

    ReplyDelete
  34. Indigo: "ottom line: will mesh adoption be slow? Sure. Are meshes DOA? No way. I still don't see a convincing argument to say that."

    I concur. I doubt meshes are DOA. Implementation will likely be very slow. Will SL still be around by that time? I have my doubts... but that's just speculation. We'll see how it comes out in the wash. : )

    (However, I'm not counting out they may be DOA. This is after all, Linden Lab. LOL)

    ReplyDelete
  35. @Wayfinder: I've made no secret of my impatience with the progress of Viewer 2 features into TPVs for a very long time, and how that's an inevitable consequence of basing on 1.x code instead of extending 2.x and fixing its many misfeatures. I'm not going to hunt for my first public post in that regard, but it probably would be in the context of Shared Media, the URI namespace, or both. But again, it's not just me: contributors to TPVs made the same points to other TPV developers, very early on. And so some TPVs based on 2.0; others didn't.

    It's quite true that all that history doesn't matter now; as I said, it's sunk cost. But good or bad, Mesh isn't going away. Viewers incapable of displaying Mesh better hope (yes "hope") that it's delayed in beta. Otherwise the users get physics envelopes that bear no relationship to what's on the screen, and this: http://www.kirstensviewer.com/Blog/66.html .

    Looking at the mesh sims on the beta grid, I don't see any reason to think all the mesh modellers are going to wait around for TPVs to catch up, and it only takes a tiny amount of Mesh content on the grid to force the point. People are not going to stumble blindly around the grid.

    ReplyDelete
  36. @Qui: No need to hunt through history for your posts; if you state you predicted an posted the need for TPVs to incorporate LL mesh, I'll take your word for it. I've seen nothing in your posts that indicates to me you're dishonorable in your statements, so your statement suffices.

    Myself, I question TPVs even considering moving to v2.x code... with the exception of dumping the interface and incorporating the new features-- with a much more sensible interface. Now that wagon I could climb on readily.

    As I stated prior, I agree that mesh probably isn't "DOA"; just like all other LL concepts, it's a matter of basic poor implementation. I don't know as I have ever seen a LL idea implemented properly and with foresight and testing before foisting it on the grid. I mean, they released Windlight and a major Havok update at the same time (I believe that was about the time of v1.16) and nearly destroyed the grid by doing so. This is not a "smart" company. So yeah, mesh will likely eventually be part of SL... but how, in what form, and when, is the major question. Possibly, as you state, even yet the TPVs will have time to implement mesh themselves. At the very least, they and the OpenSim community can get together and give LL a competitive run for its money.

    But consider: only 20% of SL is using v2.x. I doubt very many are going to switch to 2.x just so they can use mesh (some will, most won't). Will merchants go to all the trouble to develop mesh items for a board where 80% of the people can't even see them? Will you or I purchase mesh items for our land when 80% of our visitors will see either a blob or a box or a huge apple? Perhaps kicking and screaming, over a couple of years.

    I think it more likely that people will instead import pre-made mesh maps (public domain or copyrighted). It is likely there will be copyright issues over such right out of the shoot. Some folks will bring them in just because they like playing with mesh, already know how to do it, and will do it solely for their own enjoyment and the relatively few other people who will be able to see it. But I have to think that will be slow coming and slow in acceptance. It will come eventually of course, but I don't really think the TPVs need to consider themselves under any kind of supposed "deadline".

    I'll go out on a limb and make a prediction: LL will slam mesh on the main grid way ahead of its actually being ready and will cause users all kinds of problems in doing so... problems that could be prevented if they'd just exercise patience, a bit of common sense, and system design savvy.

    That's pretty much LL Mode of Operation. I don't think I'm going out very far on that prediction limb. ; )

    ReplyDelete
  37. @Wayfinder: On that prediction we agree. :)

    I'm actually no cheerleader for Mesh as it will come to Second Life. Technically, the capability is impressive, and may be tablestakes for SL's ability to compete, but it poses real risk to LL's current land-based business model, with significant reliance on a broad base of modest volume creators. That will make for interesting times ahead.

    (To be clear, I'm sure I never mentioned Mesh, per se, in my posts about TPV's choice of 1.x or 2.x codebase. I griped about nobody being able to see my MOAP maps nor use my scripts that rely on the 2.x-expanded viewer URIs. I knew about Mesh at the start of the closed beta, but even the invitation (I declined) was supposed to be confidential. There were early leaks, of course, including YouTube videos, but Mesh specifically wasn't something I was posting about at the time.)

    ReplyDelete
  38. qie, you're still assuming that there will be significant mesh uptake even in the face of unreasonable costs (potentially), a small minority of users who can see them, and significant roadblocks in the way of content creators who might want to use them. I don't think that it will work out that way at all. Regardless of how we got here, it's just not going to be that big a deal until the 60% of users who voted with their feet for TPVs can see meshes. Until then, even assuming content creators actually try to sell stuff made that way (and when your market is only 20% of what it should be, that's far from given), I doubt they will be very prevalent at all. Eye candy your friends can't see is worthless.

    ReplyDelete
  39. On the other hand, let's do some backtracking here...

    IF (and I know that's a big useless "if") LL had simply re-worked the 1.23 menus so the needed items were up front and essential items put right on the screen (such as draw distance, which we find ourselves often changing these days), IF they had added invisibility and tatoos to 1.23, IF they'd fixed the texture issue and IF they'd fixed the bugs and IF they'd decided to add meshes to 1.23... they'd have had a big, big winner.

    IF they'd have paid one of the TPV groups to update the viewer instead of a company that knew nothing about Second Life... they'd have avoided all these problems. And they'd have had a viable, functional product.

    Talk about hindsight. The shame is, that's one of the instances it should have been foresight.

    ReplyDelete
  40. I understand your arguments and they're valid,
    with one exception.

    "Meshes are also considerably less useful than they might appear to be to clothing designers, because they cannot be made flexible. A piece of rigid clothing isn't very comfortable or realistic."

    Patently untrue. Meshies can be used to create dynamic, flexible clothing, hair and accessories that are far superior to the awkward prims we have today and their accompanying mesh penetrations.

    The rest of your points are valid, but I disagree with all of them. I think that the added value that meshies bring far outweighs the downsides that you describe, and SL consumers will pick up on this very quickly.

    ReplyDelete
  41. Meshes themselves can be made to create flexible stuff - but a mesh prim in SL cannot be made flexible. This is a limitation Sl has stated publicly; the cost of making a mesh prim flexible is too high in the physics engine.

    Yes, mesh brings additional value. Where we part company is that mesh requires Viewer 2, and that is too high a hill to climb as the state of that viewer now stands.

    ReplyDelete